|
Author
|
Topic: Redundancy and degeneracy: Evolution and design
|
Frances
Member
Member # 169
|
posted 17. January 2003 12:17
In earlier threads the terms redundancy and degeneracy have been used and the ability of evolutionary mechanisms to achieve them has been questioned.
I would like to propose that we distinguish between redundancy and degeneracy.
Degeneracy: Structurally different elements can perform the same function.
redundancy: Many elements can affect the outputs in the same way but they do no thave any independent functions.
To given an example of redundancy: Airplanes have various duplicated systems that can take over a particular function if the main system fails.
The present research argues however that degeneracy is an ubiquitous biological property found at all levels of organization. It increases robustness of biological networks and adaptability to unforeseen environments.
From "Degeneracy in biological systems", Parag Ghosh, Dissertation dated 11.12.2001
An interesting recent finding is in the yeast genome which seems to have few redundant duplicate genes.
"The Yeast Protein Interaction Network Evolves Rapidly and Contains Few Redundant Duplicate Genes", Andreas Wagner, Mol. Biol. Evol. 18(7):1283–1292. 2001
And "Selection, Tinkering and Emergence in Complex Networks", Richard Sole et al
quote:
Complex biological networks have very different origins than technologic ones. The latter involve
extensive design and, as engineered structures, include a high level of optimization. The former
involve (in principle) contingency and structural constraints, with new structures being incorpo-
rated through tinkering with previously evolved modules or units. However, the observation of the
topological features of di®erent biological nets suggests that nature can have a limited repertoire
of "attractors" that essentially optimize communication under some basic constraints of cost and architecture or that allow the biological nets to reach a high degree of homeostasis. Conversely, the topological features exhibited by some technology graphs indicate that tinkering and internal constraints play a key role, in spite of the "designed" nature of these structures. Previous scenarios suggested to explain the overall trends of evolution are re-analyzed in light of topological patterns.
And of course the seminal paper "Measures of degeneracy and redundancy in biological networks", Tononi, Sporns and Edelman in Proc. Natl. Acad. Sci. Vol. 96, pp. 3257–3262, March 1999
quote:
ABSTRACT Degeneracy, the ability of elements that are structurally different to perform the same function, is a prominent property of many biological systems ranging from genes to neural networks to evolution itself. Because structurally
different elements may produce different outputs in different contexts, degeneracy should be distinguished from redundancy, which occurs when the same function is performed by identical elements. However, because of ambiguities in the distinction between structure and function and because of the lack of a theoretical treatment, these two notions often are conflated. By using information theoretical concepts, we develop here functional measures of the degeneracy and redundancy of a system with respect to a set of outputs. These measures help to distinguish the concept of degeneracy from that of redundancy and make it operationally useful. Through computer simulations of neural systems differing in connectivity, we show that degeneracy is low both for systems in which each element affects the output independently and for redundant systems in which many elements can affect the output in
a similar way but do not have independent effects. By contrast, degeneracy is high for systems in which many different elements can affect the output in a similar way and at the same
time can have independent effects. We demonstrate that networks that have been selected for degeneracy have high values of complexity, a measure of the average mutual information
between the subsets of a system. These measures
promise to be useful in characterizing and understanding the functional robustness and adaptability of biological networks.
One final paper "Degeneracy and complexity in biological systems" by Edelman and Gally in PNAS November 20, 2001 vol. 98 no. 24 13763–13768
How does the differences between redundancy (an engineering concept) and degeneracy help us determine the nature of 'design' in biological systems?
[ 17. January 2003, 12:19: Message edited by: Frances ]
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 17. January 2003 17:54
quote:
I would like to propose that we distinguish between redundancy and degeneracy.
Degeneracy: Structurally different elements can perform the same function.
redundancy: Many elements can affect the outputs in the same way but they do no thave any independent functions.
For myself, I can agree with this statement borrowed from the Abstract that Frances posted:
quote:
degeneracy should be distinguished from redundancy, which occurs when the same function is performed by identical elements.
Therefore:
Redundancy: when the same function is performed by identical elements [in the same system].
I kinda added the last part there...
quote:
Degeneracy: Structurally different elements can perform the same function.
I suppose I can agree with that, in my field we would refer to that as a compatible architecture. Just for clarification, when we are talking about the same function, are we saying "exactly" the same function? The exact same input results in the exact same output? Are the latencies equivilent? Capacities?
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 20. January 2003 23:29
My intent is to expand on the ideas of redundancy which seems to be a more stronger engineering concept and degeneracy in order to see if we can determine if biological 'design' contains redundancy or degeneracy.
Redundancy is unquestionably found in many design concepts. Backup generators, backup brakes, backup systems. But redundant systems provide for identical or similar systems to perform the same task. Redundancy however is a characteristic in which other systems can perform part or all of the function of the system in question.
Redundancy and degeneracy may be helpful concepts in determining if we can find evidence of the designers of tinkerers in biology.
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 21. January 2003 15:26
Frances,
quote:
Redundancy is unquestionably found in many design concepts. Backup generators, backup brakes, backup systems. But redundant systems provide for identical or similar systems to perform the same task. Redundancy however is a characteristic in which other systems can perform part or all of the function of the system in question.
Redundancy and degeneracy may be helpful concepts in determining if we can find evidence of the designers of tinkerers in biology.
I agree this may be a fruitfull area to investigate, but obviously we need to be very clear on our definitions and all the derivatives that may be involved. For example, a redundant system and a back-up system aren't necessarily the same thing. Let's take an aircraft. A function of a primary system is navigation. Typically, the primary system to provide the navigation function today is GPS. As a backup to that function, there is an inertial gyro system, behind that is auto direction finding using TACAN stations, behind that there may be an attitude/heading reference system, and behind that there may be a compass ball in the cockpit. All of these system are there to provide back-up services to the primary navigation system, yet each is a unique system that provides this service in a different way.
This differs from say the Space Shuttle's five identical computer systems, which then vote one their combined conclusions.
So when we talk redundancy, which are we referring to? And is that different than a "fail-over" system.
Here's a couple of other scenarios...
Many aircraft have triple redundant hydraulic systems comprised of the exact same hydraulic lines running in parallel with each other. If the primary line fails, another one is used as a backup. Now this is different than say an aircraft that has two jet engines. The aircraft is designed to fly with only one engine, so the second engine provides a "back-up" capability. However, full-functionality to achive the entire flight envelope capable (or designed into the aircraft) is only achievable with both engines running. So here we have three versions of redundancy:
Fully redundant - Dual exact systems/no performance loss
System Redundant - Different designed systems that can perform the same function with sometimes degraded performane (sometimes not).
Partially Redundant - Dual exact systems/performance loss
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 22. May 2003 23:37
I would like to explore the ideas of degeneracy and redundancy as well as how they tie in with concepts of robustness.
Any interests?
Degeneracy, Redundancy & Complexity in Biological Systems & Their Measures
Qing-jun Wang
quote:
A good portion of biological complexity comes from degeneracy. Degeneracy is used to designate different wave function satisfying the same energy state in quantum mechanics. In biology, degeneracy is “the ability of elements that are structurally different to perform the same function”. However, for many years, the concept of degeneracy is lacking and confused with redundancy, which occurs when the same function is performed by identical elements. Unlike redundant elements, degenerate elements can produce different outputs under different conditions.
Degeneracy and complexity in biological systems Gerald M. Edelman* and Joseph A. Gally
quote:
Degeneracy, the ability of elements that are structurally different to perform the same function or yield the same output, is a well known characteristic of the genetic code and immune systems.
Here, we point out that degeneracy is a ubiquitous biological property and argue that it is a feature of complexity at genetic, cellular, system, and population levels. Furthermore, it is both necessary for, and an inevitable outcome of, natural selection.
and
Selection, Tinkering, and Emergence in
Complex Networks Crossing the Land of Tinkering
[ 22. May 2003, 23:46: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 23. May 2003 19:09
Perhaps it is of interest that both concepts are relevant in linguistics. All languages have built in redundancy which aids hearing and attention. And what you're calling "degeneracy" linguists call "reanalysis" -- verbs of volition, for example, get reanalyzed as future tense, such as English will which now can be used even of inanimates, i.e., 'think it'll rain?'.
Reanalysis in language is unidirectional -- 'go/come' gets reanalyzed as 'go for a purpose' which then may get reanalyzed as future tense; allatives ('to a place') get reanalyzed as dative case ('to a conscious presence') which then might get reanalyzed as human direct object markers; 'one' and 'this' get reanalyzed as indefinite articles, 'that' gets reanalyzed as a definite article; etc. Reanalysis is a one way street -- 'the', for example, never gets reanalyzed as 'that'.
In making the analogy between linguistic reanalysis and biological "degeneracy" -- in so far as it might be an insightful analogy -- we must keep in mind that the linguistic process is one of Intelligent Design.
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 23. May 2003 19:43
Noel: In making the analogy between linguistic reanalysis and biological "degeneracy" -- in so far as it might be an insightful analogy -- we must keep in mind that the linguistic process is one of Intelligent Design.
Is it? How do we determine if linguistic reanalysis is one of intelligent design or occurred through 'mutation' and 'selection'? I am not sure if the analogy is relevant but I do not see any particular entity define how to inject degeneracy into the language.
Reanalysis
I am not sure how reanalysis can be thought to be similar to redundancy though but then again I am only an amateur linguist myself
quote:
5. Section on grammaticalization and reanalysis organised by Jan Terje Faarlund, University of Oslo (j.t.faarlund@inl.uio.no)
Invited plenary speaker: Joan Bybee, University of New Mexico.
Grammaticalization and reanalysis are two core notions in historical linguistics. Grammaticalization is usually considered a functional style gradual process, while reanalysis presupposes a formal representation of the construction involved in the change, and is thought to be abrupt or catastrophic. The way these two concepts have been used and developed in recent theories about syntactic and morphosyntactic change, it becomes more and more clear that it is not a question of two competing theories. Most functionalists recognize reanalysis as an important element in grammaticalization, while some formal linguists want to abandon grammaticalization as a separate phenomenon altogether, and rather consider it a descriptive shorthand for a set of underlying phenomena. One such phenomenon is reanalysis, which according to some may be given explanatory power, explaining various types of change, including the kinds of change that are subsumed under grammaticalization.
From my reading about reanalysis I found the following
quote:
Linguistic change is regular, and
primarily involves three factors:
acquisition
reanalysis
diffusion
Also
quote:
2. Reanalysis
Reanalysis = ‘a mechanism which changes the underlying structure of a syntactic
pattern and which does not involve any immediate or intrinsic modification of its
surface manifestation’ (Harris & Campbell 1995: 61).
Source
What fascinates me is how tools in linguistics seem to have found their usefulness in genetics and vice versa.
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 27. May 2003 16:59
Me: "... we must keep in mind that the linguistic process is one of Intelligent Design."
Pim: "Is it? How do we determine if linguistic reanalysis is one of intelligent design or occurred through 'mutation' and 'selection'?"
That's the big question! How much of language is intelligently produced by agents and how much of it is pure neurological stimulus-response? Leonard Bloomfield, a dominant figure in mid-20th century American linguistics, put it all under Skinnerian stimulus-response, only to be challenged in the later fifties by Noam Chomsky. Chomsky wished to put mind back into language, but he and his students remained firmly within the structuralist tradition (now generally called formalist). Though Chomsky does not deny the reality of mind -- of meaning and function as nonformal categories -- his syntactic component is purely formal, and consequently a phenomenon such as "reanalysis" will be described formally without reference to meaning and function.
Though not well known outside the field, Chomskian structuralism/formalism is being challenged by a growing functionalist school which sees phenomena like reanalysis/grammaticalization as something that cannot be described without reference to both structure and meaning-function. Bill Croft, for example, in a message on a functionalist discussion board, in discussing his book "Explaining Language Change: An Evolutionary Approach" (Longman, 2000), goes even further when he says, "Mechanisms of innovation are functional, mechanisms of propagation are social."
Formalists describe structure (without necessarily denying the teleological function of language), functionalists attempt to link structure and function. For functionalists a new future tense, for example, arises via metaphor (e.g., 'volition' becomes 'future', etc.), but this innovation must spread through a population and this is accomplished socially. It was William Labov who convincingly demonstrated how social factors drive change.
Some anti-ID zealots (Robert T. Pennock comes to mind) point to linguistic change in support of the Darwinian process. Then there is Dawkins' mindless "memes" (memes as ideologies unpopular in the Academy are presented as mindless, though one suspects that Darwinism itself must be a well thought out meme).
Maybe a word or two re the turbulent state of modern linguistics -- too often a world of extremes. T. Givón, one of the principle gurus of the functionalist school, complains that some of his own students are denying that language has any structure at all (see the latest edition of his Syntax), whereas Chomsky now argues a mimimalist approach (see his The Minimalist Program). The winds of postmodernism are powerful indeed.
Then there is the peculiar situation whereby the formalists, dominated by Chomsky at the far left of the political/philosophical spectrum (http://log24.com/log03/0324.htm), are often accused of being anti-Darwin, and the functionalists, with T. Givón, who argues so cogently that we take into account the teleology in language, is also a staunch Darwinist (see his Bio-Linguistics: The Santa Barbara Lectures, Benjamins, 2002). I am a student of Givón and one who sees little of use in much formalist theory, yet I find myself firmly in the camp of ID and sympathetic to Chomsky's anti-Darwinism.
[ 27. May 2003, 17:06: Message edited by: Noel Rude ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 28. May 2003 12:56
Noel: Let's try to refrain from using such words as "Zealots" for those who disagree with your ideas? This forum is making a serious attempt to avoid such rethorical arguments and it would be helpful if we can keep in mind its goals.
quote:
Zealot
1
a. One who is zealous, especially excessively so.
b. A fanatically committed person.
zealot
n : a fervent and even militant proponent of something
As far as your statement that Chomsky is anti-Darwinian, could you show us some examples of Chomsky taking such a stance?
Btw you may want to distinguish between the two following hypotheses
evolution of language
quote:
Many people have argued that the evolution of the human language faculty cannot be explained by Darwinian natural selection.
versus the evolution of _languages_ as addressed by Pennock in "Tower of Babel".
Pennock is actually quite clear in which of these hypotheses he is pursuing. Let's not confuse the two since they are very different and may lead to unnecessary confusion. One is a hypothesis about the evolution of language, the other one a hypothesis about the evolution of languages (which includes linguistic change)
As far as Chomsky and Darwinism
For a good discussion on Chomsky and Darwinism see here by Jose-Luis Mendivil Giro
quote:
Chomsky is considering the (biological) evolution of the faculty of language as a part of the evolution of human mind, but if we consider
the (historical) evolution of languages, the model of 'random historical evolution' can be applied as well, as has been shown clearly by Roger Lass in the last chapter (7) of his great book: (1997) Historical Linguistics and Language Change, Cambridge University Press, where I think you would find some answers to your questions.
[ 28. May 2003, 13:06: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 28. May 2003 16:55
Good point -- let’s not confuse language change with the evolution of language. Historical linguistics is in many ways the oldest aspect of linguistic science, but the Linguistic Society of Paris (seems to me it was sometime in the late 19th century) forbid papers dealing with the evolution of Language, a ban that was pretty much followed everywhere until somewhere in the 1970s. Everyone knew language change had nothing to do with the evolution of Language, and the theories of evolution being advanced (the “bow-wow” theory, etc.) were embarrassing.
On the one hand, as in your quote, we have the evolution of the physical brain, but language change means only change in an abstract system that can be learned by the physical brain. In the one instance we're dealing with advancement in neural complexity (specified complexity), whereas in the other there is no increase in anything beyond what the brain is already capable of doing.
However ... language change is very often compared to the Darwinian mechanism (and vice versa). This is likely because language change is a classical example of variation and selection. Yet perhaps the analogy should stop right there, because the variation in language is innovated by minds searching for structures to express various functions. Still -- there may be no better analogies of the purported power of Darwinism to create specified complexity than change in natural language and the evolution of human technology.
Both biology and (functional) linguistics study the relationship between structure and function. In linguistics we find unidirectional pathways, e.g., structural change such as
k > ch > sh > s > h > zero
lexical word > periphrasis > affix > prosody
and functional innovation which involves changes from concrete to abstract and the "bleaching" of more complex predicates. Grammatical morphemes (among the closed sets of language) come from a small number of basic verbs ('go/come', 'be', 'give', 'take', 'stay' ...) and body part words ('head', 'hand', 'front' ...).
Anyway change occurs on two levels -- concrete meanings are harnessed to perform grammatical function, and structures modify so as to make rapid ("routinized") speech more efficient.
But before we get too far off from Frances' thread … I brought this up because it seemed that linguistic reanalysis was rather like biological degeneracy ("Structurally different elements can perform the same function"). Over time a grammar might assign some function to different structures. Grammatical relations (subject, object ...), for example, can be coded by word order, affixation, intonation, ablaut -- any of the structures that are available in the vocal stream. Might comparison with reanalysis/grammaticalization help in understanding biological degeneracy? If not, then I feel uneasy about turning a good biological thread into a linguistics discussion.
You ask, "As far as your statement that Chomsky is anti-Darwinian, could you show us some examples of Chomsky taking such a stance?" Hmm ... Chomsky's writings aren't available out here in Indian country ... and anyway I'm sure Chomsky would not go on record as claiming to be anti-Darwin. But I have heard these accusations from Darwinists a number of times at linguistic conferences, and I remember concluding much the same years ago when reading Chomsky's Language and Mind. I see at Amazon (http://www.amazon.com/exec/obidos/tg/detail/-/0262032732/qid=1054048102/sr=1-10/ref=sr_1_10/104-6200645-3358337?v=glance&s=books) "Lingua ex Machina: Reconciling Darwin and Chomsky with the Human Brain," by William H. Calvin & Derek Bickerton. I took a course once from Derek Bickerton. He was a strong supporter of Chomsky's theory of innateness and also an ardent Darwinist and supporter of "punk-eek".
[ 28. May 2003, 16:59: Message edited by: Noel Rude ]
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 28. May 2003 18:17
Chomsky is typically read as putting forth a less-than-Darwinian thesis when it comes to the evolution of language. He, nonetheless, would never use this as an argument against Darwinian evolution. This reading of Chomsky has been raised in various independent contexts in several of my graduate level philosophy classes (Wittgenstein, Philosophy of Science, Philosophy of Mind). I also recall reading about it in a book by John Heil on the Philosophy of Mind.
I'll see if I can track down a friend of mine who is a Chomsky fanatic to see if he'll give us some more details.
[ 28. May 2003, 18:17: Message edited by: Micah Sparacio ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 28. May 2003 23:02
I agree that some may have refered to Chomsky as a 'crypto-creationist'
Brian Chandler
quote:
This gives him opportunities to get his teeth into a number of distinguished figures, including Stephen Jay Gould, Noam Chomsky, and Roger Penrose: the first two turn out to be (my assessment) crypto-mystics, and the last simply muddled.
also
quote:
Chomsky may be right to believe that language did not gradually evolve, but even he would admit that this idea is only speculation. He just prefers to think of the language organ as a self-enclosed system whose origins are mysterious. It is not for nothing that he has been called a "crypto-creationist." Steven Pinker, an admirer of both Chomsky and Darwin, thinks that Chomsky's distaste stems from a more general dislike of arguments that derive human qualities from utility. The theory of natural selection, after all, assumes that things evolve because they are useful; in that, it is a larger version of the behaviorist thesis that humans, like animals, do things in order to get stuff for themselves. And it is true that Chomsky believes that humans are driven by the desire for creative expression, not by anything so crass and petty as advantage. Daniel Dennett, a philosopher at Tufts University, believes that Chomsky's resistance is also due to a dislike of the ad-hoc, gadgetry aspect of evolution: Chomsky wants to think of language as a perfect, unified system....
source
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 29. May 2003 08:48
The question, though, Pim, is whether Chomsky has any good reasons for disliking Darwinian explanations for the origin of language...
The crypto-mystic stuff is typical Dennett. Either someone is fully consumed by the universal acid or they're on Dennett's hit list. It seems that there really is *no* good reason to question Darwinian explanations, from a Dennettian point of view.
It seems to me that Chomsky is just passionate about language and through his research has come to see the language organ, in many ways, as an IC system (though he wouldn't call it that).
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 29. May 2003 12:26
Micah: The question, though, Pim, is whether Chomsky has any good reasons for disliking Darwinian explanations for the origin of language...
Indeed, which is why I was somewhat surprised to see Chomsky refered to as anti-Darwinian.
Micah: It seems to me that Chomsky is just passionate about language and through his research has come to see the language organ, in many ways, as an IC system (though he wouldn't call it that).
ICness is not a reliable indicator of ID though. However Pennock's comparison of biological and linguistic evolution did not look at the evolution of language but rather evolution of languages.
quote:
"A rather large irony overshadows Steven Pinker's career-and all of evolutionary psychology. Pinker in particular and evolutionary psychologists in general are deeply indebted to the linguist Noam Chomsky. Chomsky, who like Pinker is an MIT professor, pioneered the genetic, modular approach to the mind pursued by evolutionary psychologists. ... when I discussed the issue with him, Chomsky insisted that his doubts about neo-Darwinism are purely scientific. He accepted that natural selection probably played some role in the evolution of language
and other human attributes. But given the enormous gap between human cognitive capacities and those of other animals, he thought that science could say little about how or why those capacities evolved. Darwin's theory essentially says that there is `a naturalistic explanation for things,' Chomsky elaborated. Anyone who does not believe in `divine intervention' accepts as
much. The difficulty lies in determining what the correct naturalistic explanation *is*. Natural selection is `*a* factor in determining the
distribution of traits and properties within these constraints. *A* factor, not *the* factor.'" (Horgan J., "The Undiscovered Mind: How the Brain Defies Explanation," [1999], Phoenix: London, 2000, pp.177-178. Emphasis in original)
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 29. May 2003 13:02
Whoa! First off, I was only referring to a "less-than-Darwinian" (not the same as an anti-Darwinian) Chomsky. Second, he would never view himself in this way... the context here is that many scholars have read him this way.
Finally, I just contacted my friend who is a Chomsky buff, and he thinks that to say that Chomsky is non-Darwinian is a bit too simplistic. Rather, Chomsky thinks that human knowledge doesn't come easily, and he is opposed to explanations that fail the task of increasing human knowledge. Unfortunately, many Darwinian explanations miss this threshold of explanatory specification (the universal acid that explains everything) ... though it certainly doesn't preclude good, solid Darwinian explanations.
Basically, Chomsky is merely a stickler when it comes to explanation, and especially when it comes to his discipline, linguistics, he prefers that the explanations be more than off-the cuff generalizations or speculations.
[ 29. May 2003, 13:03: Message edited by: Micah Sparacio ]
IP: Logged
|
|
|
Printer-friendly view of this topic